Review Article

Fluoride in Drinking Water and Nanotechnological Approaches for Eliminating Excess Fluoride

Table 1

Summary of fluoride removal capacities by different adsorbents.

NoAdsorbantSpecial remarksAC (mg/g)Studies conductedRef

1MgO-MgFe2O3 binary oxide anchored on to GO (graphene oxide)Adsorption mechanism:
Electrostatic interactions inner-sphere complexation (Mg2+ -F) (anion exchange)
H bonding hydrothermal method (precursor; PEG)
Ave. diameter of binary oxide: 10 nm separation via an external magnetic field
Highly stable and high specific area presence of CO32− and PO43− lowers the removal %
34 at pH 6.0 (303.15 K)Pseudo-second-order kinetic equation
Langmuir isotherm model
Endothermic and spontaneous process
Saturation magnetization:
0.575 emg−1
[71]

2rGO/ZrO2 (continuous fixed bed column study in upward flow mode)Hydrothermal route
Possible mechanisms: Surface reactions between ZrO2 and F(ZrO2 exists as seven coordinated polyhedral species and hence it increases the adsorption capacity of rGO)
Addition of F happens via hydroxyl protonation
Fluoride ion attack loss of water molecule
Removal was affected by PO43−
45.7 at pH 7Break through curve: Yoon–Nelson model (continuous bed type)[95]

33D yttrium based-graphene oxide-sodium alginate hydrogelMacro structure adsorbent which involves graphene oxide. A sol-gel process was used
GO sheets can be easily crosslinked with yttrium ion and sodium gel was used as the supporter medium.
Mechanism: ion exchange
Easy regeneration
288.96 at pH 4.0Langmuir isotherm model
Both pseudo-first-order and pseudo-second-order models are best matched.
Column study:
Thomas model. The removal rate in the column system (at relatively lower flow rate) >batch reactor
[96]

4Hydrous iron (iii)-aluminum (iii) mixed oxide-graphene oxide composite (HIAGO)Synthesized by a chemical precipitation method
“Spindle liked” mixed oxides dispersed irregularly on GO backbone
Mechanisms: Electrostatic interaction and ion exchange
27.8 at pH 5 (318 K)Best matched with Langmuir isotherm model and the pseudo-second-order kinetics endothermic reaction[94]

5Hierarchical AlOOH@ reduced graphene oxide hybridSupercritical carbon dioxide assisted synthesis (eco-friendly)
Specific area: 513 m2·g−1 (mesoporous structure)
AlOOH nanosheets were evenly spread (layer upon layer) on rGO and a hierarchical 3D structure was created
Electrostatic interaction and ligand exchange adsorption process to form Al-F complexes
Mean pore size; 8 nm
118.7 at pH 6.5 (high AC at lower F concentration)Best matched with Langmuir isotherm model and the pseudo-second-order kinetics[97]

6Magnetic iron-aluminum oxide/graphene oxide nanoparticlesOne step method of co -precipitation method
Specific surface area; 349 m2/g
Mechanisms: Electrostatic interactions, anion exchange, and inner-sphere complexation
Can be used in a wide pH range; 3–9 low residual ion and aluminum residual after defluoridation
64.72 at pH 6.5Best matched with Langmuir isotherm model and the pseudo-second-order kinetics. Saturation magnetization:
7.5 emu/g (superparamagnetism)
[98]

7Al2O3-Fe3O4-expanded graphite nano-sandwich structureCoprecipitate method resulting 60–80 nm Fe3O4 cubes and 20–50 nm Al2O3 spread on graphite surface
Fluoride removal efficiency:
96% at pH 4
After two cycles the efficiency: 91%
Convenient separation via an external magnetic field Stability is higher
2.19 at pH 2–10Best matched with Langmuir isotherm model and the pseudo-second-order kinetics. Saturation magnetization:
3.7 emu/g
[99]

8Hydrous CeO2-Fe3O4 decorated polyaniline fibers nanocompositeA coprecipitation deposition on pre-synthesized polyaniline fibers over a large pH range (3–10) the removal rate was nearly constant giving the maximum as 94% at pH 393.46–117.64 over a pH range 3–10Best matched with Langmuir isotherm model and the pseudo-second-order kinetics. Exothermic reaction. Spontaneous reaction[100]

9Activated carbon@SnO2 (biosorption process)Activated carbon was synthesized via pyrolysis of sawdust and activation with phosphoric acid. A mixture of activated C and stannous chloride was sonicated followed by NH3 addition to obtain a gel like product. After washing the obtained product was calcined at 673 K for 2 hCrystal structure; rutile tetragonal tin oxide SnO2 nanoparticles embedded on amorphous carbon matrix (12 nm)
Mechanism; combination of strong electrostatic interaction between Sn4+ and arsenic/fluoride anions, followed by anion exchange process with hydroxylated surface of SnO2-AC nanocomposite
4.60 at pH 6.5Best matched with Langmuir isotherm model and the pseudo-first-order kinetics. Endothermic and spontaneous reaction. Chemisorption may happen. Greater regeneration capability (removal % remained around 90% after 3 consecutive cycles)
Low production cost (cheaper than activated C solely according to the cost estimation study)
[101]

10Magnesium oxide entrapped polypyrrole hybrid nanocompositeMgO nanoparticles were synthesized with constant mechanical stirring (1000 rpm) using MgCl2 and NH3. Polymer nanocomposite was synthesized by chemical oxidative polymerization using FeCl3 as an oxidant.4.32Best matched with Langmuir isotherm model fluoride removal is endothermic and spontaneous nature.
Field study was conducted. (79% of fluoride was removed)
[102]

11Mesoporous γ-ALOOH powder from coal-bearing kaoliniteExtraction of Al from kaolinite using sulfuric acid instead of hydrochloric acid. Synthesis of mesoporous γAlOOH was performed in a hydrothermal synthesis using hexamethylenetetramine as the hydrolyzing agent
The average crystallite size was estimates as 14.4 nm for as prepared material compared to commercial one in which the value is 37 nm.
Not givenBET surface area could be controlled from 6.3 to 192.5 m2·g−1 when varying reaction time and temperature.[103]

12Chitosan-Fe-Al-Mn metal oxyhydroxides compositeHydrous mixed-metal oxyhydroxides, loaded chitosan composite was made using laterite clay and waste from steel (low-cost materials) industry via coprecipitation method
Polymeric matrix prevents particle agglomeration
20–30 nm mixed-metal spheres were homogeneously spread on chitosan matrix (this ensures easy access of F-onto the porous outer surface of mixed-metal sphere) BET surface area is 41.6 m2/g. Influence of other ions was not significant.
Prevents strong electric double layer formation when the polymer barrier is present and thus it does not hinder the adsorption. can be used as a fluoride scavenging material in a household filter effective for broad pH range; 3.5–8.5
40.0 ± 0.5 at pH 6.7Best matched with Freundlich isotherm model (multilayer adsorption on heterogenous surface sites) and the pseudo-second-order kinetics with rapid adsorption (chemisorption is possible)
Adsorption capacity of mixed-metal oxyhydroxides with chitosan > without chitosan
[104]

13Highly efficient nano- adsorbent, Al(III)-Fe(III)-La(III) trimetallic oxideHealth risk of using Al solely has been eliminated by introducing Fe3+ in this study (although Fe3+ is widely used for fluoride adsorption some studies suggest low AC as 16.5 mg/g). La3+ is an excellent fluoride adsorbent. The presence of Fe3+ makes the separation easier via an external magnet
Chemical rout was used and triethanolamine is the chelating agent
Adsorption mechanism: Ligand exchange
XRD pattern revealed the amorphousness of the structure which increase the surface area (22.14 m2/g)Avrg. Particle size; 20 (±0.50) nm pore diameter 3.324 nm
Not givenMaximum fluoride could be removed up to 99.8% at pH 7.0, contact time: 60 min, adsorbent dose: 0.3 g/100 ml[105]

14Hybrid Al2O3/Bio-TiO2 nanocomposite impregnated
Thermoplastic polyurethane (TPU) nanofibrous membrane for fluoride removal
Titanium dioxide was synthesized using a bacteria Bacillus licheniformis and modified with alumina followed by impregnation of synthesized hybrid material on to eletrospun thermoplastic polyurethane nanofibers through silane functionalization
Particle size of the nanocomposite: 50 ± 6 nm. Average diameter: 239 ± 33 nm. Titania was in the anatase phase having a body centered tetragonal crystal structure. Mechanism of F adsorption: Formation of surface complexes
1.9Both batch type and dip mode adsorption studies were carried out and, in both studies, adsorbent capacity decreased when the adsorbate concentration is increased.
Best matched with Langmuir isotherm model and with second-order kinetic model.
[106]

15Coconut-shell derived carbon/carbon nanotube compositeCNT were coated on coconut-shell charcoal using plasma enhanced chemical vapor deposition
BET surface area: 358 m2·g−1
Low-cost adsorbents Removal percentage: 65% of the initial concentration of 4.4 mg L−1
0.36 at pH 2Best matched with Langmuir isotherm model and the pseudo-second-order kinetics fluoride removal is endothermic and spontaneous nature. A field study was carried out to nalagonda water sample and obtained satisfactory results (initial adsorbent concentration is 10 g L−1 and equilibrium contact time was 3.5 h)[107]

16Iron-oxide nanoparticles by green Synthesis. Method using Moringa oleifera leaf extractGreen synthesis (a solution of iron nitrate III was mixed with heated leaf extract under stirring and immediate color change from green to black indicates the nanoparticle formation)
Plant extract acts as the reducing agent
Pore diameter: 4.14 nm. Specific surface area: 99.79 m2·g−1
Spherical mesopores were present. There is no need to use chemicals, high temperatures, and pressure
1.40 at pH 7 (higher than BGAC)Study compared the removal percentage with granulated charcoal of bone (BGAC)
Best matched with Langmuir isotherm model and the pseudo-first-order kinetics. Endothermic and spontaneous. Equilibrium time: 40 min at pH 7 for iron oxide while it was 180 min at pH 5for BGAC
Repeated sorption regeneration cycles indicated reusability.
[108]

17NaP : HAp nanocompositeHydroxyapatite was synthesized followed by composite making with NaP Zeolite
Both cactus-like and diamond like morphology were observed
Particles shaped hydroxyapatite were within the diameter of 40–70 nm
BET surface area: 35.62 m2/g pore volume: 0.141 cm3/g
66.66 mg/gAnalysis of several factors were carried out using the four factors box—Behnken design with three levels (optimum pH value was found to be 7 at 60 min with 5 mg/L of fluoride concentration and at 55°C)
Optimum removal; 97.45%
Best matched with Langmuir isotherm model. The adsorption is an endothermic process
[109]

18Millisphere nanocomposite of La-Doped Li-Al layered double hydroxides supported by polymeric anion exchanger (LaLiAl-LDH@201)Drawbacks of LDHs such as pressure drop and blockage were overcome by impregnating the LDHs into millimetric polystyrene anion exchanger
Li/Al LDH was doped with high-valent metal Lanthanum
Spherical nanoparticles with clean edges to irregular state.
Stability over wide pH range; 3.5–12
Adsorption mechanism: Anion exchange between Cl− and F−, and the specific interaction between La and F
Capability of in situ regeneration when NaOH, NaCl binary mixture was added
An increase in adsorptive capacity in the presence of the competing anions was observed
Negligible capacity loss, La leaching, or structure alteration was observed after five adsorption-regeneration cycles
75.7The adsorption kinetics could be fitted with both pseudo-first-order and pseudo-second-order models with approximately identical coefficients of determination.
Best matched with Freundlich isotherm model. Excellent fixed bed working capacity and regeneration ability.
[110]

19Tetraethylenep-entamine functionalized nanomagnetic compositenFe3O4@TEPA was produced using a polyol-media one-pot solvothermal method. Adsorption capacity was independent from solution pH (2–11). The effect of HF at pH below 5 (pHpzc) has not affected the AC. Avg. diameter: 20 nm. Higher saturation magnetization compared to other nanomagnetic polymers.
Adsorption mechanism: Electrostatic attraction and H bonding. The effect of coexisting ions were estimated by standard response surface methodology (RSM)design called box-behnken design (BBD)
163.9Best matched with Langmuir isotherm model. Endothermic and spontaneous saturation magnetization: 48.2 emu/g. Better reusability (composite could be used for at least 10 cycles with a loss of less than 2.8% upon recovery on average)[111]

20Membrane capacitive deionization with nanoporous and microporous Limonia acidissima (wood apple) shell activated carbon electrode (LASAC)Nanoporous-activated C was synthesized from Limonia acidissima using chemical and thermal modifications. For LASAC electrode preparation, LASAC powder, graphite powder, and polyvinylidene fluoride were mechanically stirred in N, N dimethylacetamide solvent to make a paste2.7554 at pH 7.2
Under 100 mg/L
Feed concentration and 1.2 V
Best matched with Langmuir isotherm model and the pseudo-first-order kinetics[112]

21Electrically enhanced adsorption and green regeneration 1 for fluoride
2 removal using Ti(OH)4-loaded activated carbon electrodes
Blending of electrosorption and adsorbent adsorption
Membrane can be easily regenerated by applying -1.6 V
Excellent regeneration capability and stability of removal % after reusage
The adsorbent possesses high adsorption F selectivity, high adsorption capacity, and clean regeneration without any chemicals
Surface area: 1700 m2/g
Pore size: 1.5 nm and 2.35 nm
Influence of other ions is negligible
115.2 at +1.2 VBest matched with Langmuir isotherm model[113]

22Millimeter-sized Mg-Al-LDH nanoflake impregnated magnetic alginate beadsBiobased sorbent
Rhombohedral symmetry was there. Estimated thickness of the nano-LDH: 18 nm
Multidispersed Fe2O3 particles with avg. diameter 20 nm surface adsorption, interparticle diffusion and intraparticle diffusion all contributing to the rate of adsorption
32.4 at pH 5Best matched with Freundlich isotherm model and the pseudo-second-order kinetics with rapid adsorption[57]

23Metal organic framework (Uio-66)ZrCl4 [Zr6O4(OH)4(BDC)6]21 and terephthalic acid were dissolved in DMF solution and heated in an autoclave to synthesis Uio-66
Avg. size: 153–213 nm. Optimum conditions according to RSM and central composite design analysis were pH: 7, fluoride ion concentration: 14.6 mg L−1, Uio-66 dosage: 0.4 g l–1, contact time: 41.5 minute, removal efficiency of 80.21 percent, and a desirability of 1
31.09 pH range 6–9Best matched with Langmuir isotherm model and the pseudo-second-order kinetics.[114]